In general, when viewing a movie or an advertisement program on a television (TV) and the like, when finding an object (for example, a person or a thing) that a user favors while shopping, or when finding a predetermined object that a user is usually interested in, users desire to memorize the object by taking a memo of the object or currently by capturing the object using a camera of a mobile terminal.
However, in the first case, management is not easy in that a user needs to make a habit of taking a memo, and to carry a memo pad at all times, and also may lose a memo pad in which detailed information of a product is written.
In the second case, portability and management is easy in that a user uses a mobile terminal that the user carries all the time. However, there is a disadvantage in that the user needs to manually input information exposed in an image in order to search for a desired object using the captured image. For example, the user needs to input a name or a product number of a product exposed in the image, a shape, a color, and the like.
Here, with respect to clearly exposed information such as the name or the product number of the product, a search may be clearly performed. However, when there is a characteristic in the shape, the image, and the like, or when accurate information is unknown even after viewing the image, it is difficult to perform a search.
Accordingly, currently, technologies embedded with a decoder (a recognition engine) capable of performing recognition using an image of an object have been applied to a terminal.
That is, disclosed is technology of recognizing a barcode from a barcode image by equipping a terminal with a barcode decoder, or recognizing a quick response (QR) code from an image of the QR code using a terminal embedded with a QR code decoder.
However, in the above technologies, decoders for a barcode, a QR code, facial recognition, and image recognition are individually equipped in each terminal and thus, recognition is limited according to a type of an object. Also, to recognize an object, there is disadvantage in that a user needs to directly execute a decoder according to the type of the object.